Flavor Tomography Alessandro Buzzatti Miklos Gyulassy April 17

Flavor Tomography Alessandro Buzzatti Miklos Gyulassy April 17 th, 2013 Alessandro Buzzatti – LBNL 1

Outline • Introduction – Jet tomography and jet quenching • CUJET – Analysis of the model – Geometry, path integral, partonic spectra • Jet tomography at RHIC and LHC – Heavy quark puzzle (RHIC) – Surprising transparency (LHC) • Future developments – Elliptic flow • Conclusions April 17 th, 2013 Alessandro Buzzatti – LBNL 2

Quark Gluon Plasma Perturbative QCD Hydrodynamics • Medium is strongly coupled • Local thermal equilibrium is maintained • Hydrodynamic expansion and collective flow • Measurement of ‘bulk’ properties of QGP April 17 th, 2013 • Alessandro Buzzatti – LBNL 3

Jet Tomography g g q, g π, η g Measurement of hadronic quenching pattern provides information about QGP density • Requires detailed knowledge of jet-medium interaction mechanism and jet transverse diffusion • Knowledge of initial conditions, `cold’ nuclear matter effects and hadronization process April 17 th, 2013 Alessandro Buzzatti – LBNL 4

RAA measurements Cold nuclear effects Jet quenching April 17 th, 2013 Alessandro Buzzatti – LBNL 5

Jet quenching • April 17 th, 2013 Alessandro Buzzatti – LBNL 6

Model comparison (Opacity expansion) (Multiple soft scattering) Horowitz, HP 12 April 17 th, 2013 Alessandro Buzzatti – LBNL 7

Outline • Introduction – Jet tomography and jet quenching • CUJET – Analysis of the model – Geometry, path integral, partonic spectra • Jet tomography at RHIC and LHC – Heavy quark puzzle (RHIC) – Surprising transparency (LHC) • Future developments – Elliptic flow • Conclusions April 17 th, 2013 Alessandro Buzzatti – LBNL 8

History • Genealogy of Jet tomography at Columbia University: Vitev, Molnar, Djordjevic, Adil, Wicks, Horowitz, Ficnar. • Past efforts culminated with the construction of the WHDG model – – Radiative (static potential) + Elastic energy loss Static medium (approximated at half the time of the expansion) Static jet-medium coupling parameters Simplified convolution over partonic spectra (spectral index approximation) • Applications of WHDG to RHIC and LHC phenomenology got mixed response – Correct predictions of flat pion RAA behavior at RHIC – Under-quenching of non-photonic electrons at RHIC – Over-quenching of pion RAA at LHC April 17 th, 2013 Alessandro Buzzatti – LBNL 9

CUJET • Geometry – Glauber model – Bjorken longitudinal expansion • Energy loss – Full jet path length integration – Collisional energy losses – Gluon emission fluctuation effects • Detailed convolution over initial production spectra • In vacuum Fragmentation Functions April 17 th, 2013 Alessandro Buzzatti – LBNL 10

Bjorken expansion • MONOTONIC density dependence April 17 th, 2013 Alessandro Buzzatti – LBNL 11

CUJET • Geometry – Glauber model – Bjorken longitudinal expansion • Energy loss – Full jet path length integration Possibility to evaluate systematic theoretical uncertainties such as sensitivity to formation and decoupling phases of the QGP evolution, local running coupling and screening scale variations, and other effects out of reach with analytic approximations. – Collisional energy losses – Gluon emission fluctuation effects • Detailed convolution over initial production spectra • In vacuum Fragmentation Functions April 17 th, 2013 Alessandro Buzzatti – LBNL 12

Strong running coupling • Introduce one-loop alpha running B. G. Zakharov, JETP Lett. 88 (2008) 781 -786 – – S. Peigne and A. Peshier, Phys. Rev. D 77 (2008) 114017 April 17 th, 2013 Alessandro Buzzatti – LBNL 13

CUJET • Geometry – Glauber model – Bjorken longitudinal expansion • Energy loss – Full jet path length integration Possibility to evaluate systematic theoretical uncertainties such as sensitivity to formation and decoupling phases of the QGP evolution, local running coupling and screening scale variations, and other effects out of reach with analytic approximations. – Collisional energy losses – Gluon emission fluctuation effects • Detailed convolution over initial production spectra • In vacuum Fragmentation Functions April 17 th, 2013 Alessandro Buzzatti – LBNL 14

Production spectra RHIC LHC NLO-FONLL uncertainty April 17 th, 2013 Vogt, Xin-Nian Alessandro Buzzatti – LBNL 15

CUJET • Geometry – Glauber model – Bjorken longitudinal expansion • Energy loss – Full jet path length integration Possibility to evaluate systematic theoretical uncertainties such as sensitivity to formation and decoupling phases of the QGP evolution, local running coupling and screening scale variations, and other effects out of reach with analytic approximations. – Collisional energy losses – Gluon emission fluctuation effects • Detailed convolution over initial production spectra • In vacuum Fragmentation Functions April 17 th, 2013 Alessandro Buzzatti – LBNL 16

Systematic errors • Initial conditions and plasma evolution – Thermalization phase, plasma density profile and fluctuations, plasma constituents, hydrodynamic expansion, freeze-out temperature • pp reference spectra • Jet-medium coupling – Interaction potential, order of opacity expansion, running coupling scales, multiple gluon emission, large angle radiation April 17 th, 2013 Alessandro Buzzatti – LBNL 17

Running coupling • a-b-c model Running scales Gyulassy, Horowitz GLV fixed all LHC Ø Increase in the slope of RAA over a broad range of momenta April 17 th, 2013 Alessandro Buzzatti – LBNL 18

Initial p. QCD spectra Initial quark production spectra RHIC GLUE UP CHARM BOTTOM April 17 th, 2013 Competing effects between increased density and harder production spectra – RHIC density and spectra – LHC density, RHIC spectra – LHC density and spectra UP BOTTOM Alessandro Buzzatti – LBNL 19

Initial p. QCD spectra Initial quark production spectra RHIC GLUE UP CHARM BOTTOM April 17 th, 2013 Competing effects between increased density and harder production spectra – RHIC density and spectra – LHC density, RHIC spectra – LHC density and spectra UP BOTTOM Alessandro Buzzatti – LBNL 20

Initial p. QCD spectra Initial quark production spectra LHC GLUE UP CHARM BOTTOM April 17 th, 2013 Competing effects between increased density and harder production spectra – RHIC density and spectra – LHC density, RHIC spectra – LHC density and spectra UP BOTTOM Alessandro Buzzatti – LBNL 21

Outline • Introduction – Jet tomography and jet quenching • CUJET – Analysis of the model – Geometry, path integral, partonic spectra • Jet tomography at RHIC and LHC – Heavy quark puzzle (RHIC) – Surprising transparency (LHC) • Future developments – Elliptic flow • Conclusions April 17 th, 2013 Alessandro Buzzatti – LBNL 22

Explained issues Heavy quark puzzle (RHIC) April 17 th, 2013 Surprising transparency (LHC) Alessandro Buzzatti – LBNL 23

Level crossing at RHIC fit Glauber+Bjorken, dynamical, rad+col central collision d. N/dy=1000 fixed coupling, αs=0. 3 τ0=1 fm/c b u c g April 17 th, 2013 Alessandro Buzzatti – LBNL 24

RHIC data RHIC fit Glauber+Bjorken, dynamical, rad+col central collision d. N/dy=1000 fixed coupling, αs=0. 3 CUJET @ RHIC ELECTRONS (PHENIX) PIONS (PHENIX) Ø Reduced discrepancy between pions and non-photonic electrons April 17 th, 2013 Alessandro Buzzatti – LBNL 25

Level crossing at LHC extrapolation parameter-free d. N/dy=2200 LHC production spectra b u c g April 17 th, 2013 Alessandro Buzzatti – LBNL 26

LHC data LHC extrapolation parameter-free (fixed coupling, αs=0. 3) d. N/dy=2200 LHC production spectra CUJET @ LHC PIONS (CMS) ELECTRONS (PHENIX) Ø Predictions are over-quenched: the QGP at LHC is surprisingly transparent to hard probes Ø Possibility of reduced coupling? PIONS (ALICE) PIONS (PHENIX) April 17 th, 2013 Alessandro Buzzatti – LBNL 27

LHC data with running coupling LHC fit, RHIC extrapolation parameter-free Glauber+Bjorken, dynamical, rad+col central collision d. N/dy=2200 running coupling, α 0=0. 4 CUJET @ LHC PIONS (PHENIX) PIONS (CMS) PIONS (ALICE) Ø Surprising agreement with data Ø Consistent with backward extrapolated RHIC predictions April 17 th, 2013 Alessandro Buzzatti – LBNL 28

Open heavy flavor data CUJET @ LHC fit, RHIC extrapolation parameter-free Constrained by LHC pion fit PIONS (PHENIX) ELECTRONS (ALICE) PIONS (CMS) PIONS (ALICE) D mesons (ALICE) April 17 th, 2013 Alessandro Buzzatti – LBNL 29

Open heavy flavor data CUJET @ LHC fit, RHIC extrapolation parameter-free Constrained by LHC pion fit PIONS (PHENIX) PIONS (CMS) PIONS (ALICE) April 17 th, 2013 Alessandro Buzzatti – LBNL 30

Outline • Introduction – Jet tomography and jet quenching • CUJET – Analysis of the model – Geometry, path integral, partonic spectra • Jet tomography at RHIC and LHC – Heavy quark puzzle (RHIC) – Surprising transparency (LHC) • Future developments – Elliptic flow • Conclusions April 17 th, 2013 Alessandro Buzzatti – LBNL 31

Elliptic flow April 17 th, 2013 Alessandro Buzzatti – LBNL 32

Elliptic flow April 17 th, 2013 Alessandro Buzzatti – LBNL 33

Hydro expansion April 17 th, 2013 Alessandro Buzzatti – LBNL 34

More on hydro expansion April 17 th, 2013 Alessandro Buzzatti – LBNL 35

Conclusions • Quark Gluon Plasma – Low- hydro vs. high- p. QCD – QGP properties can be studied measuring high- Jets – Jet tomography requires knowledge of: (1) initial conditions (2) plasma evolution (3) energy loss (4) jet fragmentation • CUJET – Quantitative predictions insensitive to many of the systematic theoretical uncertainties – Evolution over WHDG » Geometry, Jet path integration, Partonic spectra, Dynamical potential – Predicted level crossing pattern, reduced light-bottom quark jets discrepancy, steep rise of RAA at LHC • Future – Address azimuthal asymmetry puzzle » Include viscous hydrodynamics – Full jet measurements April 17 th, 2013 Alessandro Buzzatti – LBNL 36

WHDG • Elliptic flow Ø Need stronger coupling between jet and medium, but RAA will be over-quenched Ø The agreement between data and theoretical curves worsen for higher centralities and lower transverse momenta. April 17 th, 2013 Alessandro Buzzatti – LBNL 37

Elliptic flow April 17 th, 2013 Alessandro Buzzatti – LBNL 38

Radiative energy loss Incoherent limit: Gunion-Bertsch • Formation time physics • April 17 th, 2013 Alessandro Buzzatti – LBNL 39

GLV interference effects • Energy loss LPM interference effects with hard production vertex (jet created at finite time) • Transverse diffusion LPM interference effects with successive scatterings (gluon cascading) April 17 th, 2013 Alessandro Buzzatti – LBNL 40

DGLV mass effects April 17 th, 2013 Alessandro Buzzatti – LBNL 41

Effective Potential Static potential (DGLV) Dynamical potential (MD) • Static scattering centers • Color-electric screened Yukawa potential (Debye mass) • Full opacity series • Dynamical scattering centers • Includes screened color-magnetic effects (HTL gluon propagators) • Only first order in opacity Interpolating potential (CUJET) • Introduces effective Debye magnetic mass • Interpolates between the static and dynamical limits • Magnetic screening allows full opacity series April 17 th, 2013 Alessandro Buzzatti – LBNL 42

Outline • Introduction – – QCD phase diagrams and equation of state Heavy ion collisions and formation of Quark Gluon Plasma Jet tomography and jet quenching Energy loss models • Opacity expansion series – Convergence of the series – Radiated gluon transverse momentum distribution • CUJET – Analysis of the model – Geometry, path integral, partonic spectra • Jet tomography at RHIC and LHC – Heavy quark puzzle (RHIC) – Surprising transparency (LHC) • Conclusions April 17 th, 2013 Alessandro Buzzatti – LBNL 43

Convergence of the series Static brick E=50 Ge. V ω=5 Ge. V (x=0. 1) L=5 fm T=250 Me. V Ø Interfering effects dominated by n=1 order in opacity Ø Convergence already at n=5 April 17 th, 2013 Alessandro Buzzatti – LBNL 44

Light-charm quark similarity Light (dashed) and charm (solid) Bottom Static brick E=20 Ge. V ω=5 Ge. V (x=0. 25) Ml=0. 2 Ge. V Mc=1. 2 Ge. V Mb=4. 75 Ge. V L=5 fm T=250 Me. V (vacuum rad) Dead cone Ø Charm and light quark transverse momentum spectra are almost identical Ø Bottom quark radiation is suppressed by heavy mass Ø Dead cone filled by induced radiation April 17 th, 2013 Alessandro Buzzatti – LBNL 45

‘Thin’-‘thick’ plasma approximation Static brick E=20 Ge. V ω=5 Ge. V (x=0. 25) L=5 fm T=250 Me. V (multiple soft scattering) April 17 th, 2013 Alessandro Buzzatti – LBNL 46

n=1 approximation Light to heavy quark energy loss ratio Static brick E=20 Ge. V ω=[integrated] Ml=0. 2 Ge. V Mh=4. 75 Ge. V L=5 fm T=250 Me. V Ø The energy loss ratio between light and heavy quark jets is insensitive to higher order in opacity corrections Ø Flavor tomography can be studied at n=1 Ø Jet measurements require full transverse radiative spectrum at high orders April 17 th, 2013 Alessandro Buzzatti – LBNL 47

Interpolating potential static dynamical Static brick E=10 Ge. V ω=[integrated] T=250 Me. V Ø Strong indication of similar convergence for static and dynamical potentials April 17 th, 2013 Alessandro Buzzatti – LBNL 48

Static and dynamical limits Static (dashed) and dynamical (solid) energy loss ΔE/E Static brick E=20 Ge. V Ml=0. 2 Ge. V Mc=1. 2 Ge. V Mb=4. 75 Ge. V L=5 fm T=250 Me. V Ø The dynamical potential increases the energy loss Ø Quadratic L-dependence for light and charm (LPM effect) vs. linear for bottom Ø Remarkably different dependence on E for bottom April 17 th, 2013 Alessandro Buzzatti – LBNL 49

Thermalization RHIC LHC Glauber+Bjorken d. N/dy=1000 (RHIC) 2200 (LHC) τ0=1 fm/c dynamical potential fixed coupling radiative+collisional Ø Large sensitivity to pre-thermalization phase, canceled by a small (± 10%) rescaling of the coupling constant April 17 th, 2013 Alessandro Buzzatti – LBNL 50

BDMPS April 17 th, 2013 Alessandro Buzzatti – LBNL 51

ASW April 17 th, 2013 Alessandro Buzzatti – LBNL 52

Dynamical potential Aurenche, Gelis, et al, 2002 Pisarski, 1980 Djordjevic, Heinz, 2007 April 17 th, 2013 Alessandro Buzzatti – LBNL 53

Gunion-Bertsch April 17 th, 2013 Alessandro Buzzatti – LBNL 54

Elastic energy loss and Fluctuations Bjorken elastic collisions • Energy loss fluctuations • Poisson expansion of the number of INCOHERENTLY emitted gluons Gaussian fluctuations April 17 th, 2013 Alessandro Buzzatti – LBNL 55

Geometry of the collision Glauber model is used to compute number of binary collisions and participant nucleons • April 17 th, 2013 Au. Au Alessandro Buzzatti – LBNL 56

k. T sensitivity • 20 Ge. V, L = 5, light quark April 17 th, 2013 Alessandro Buzzatti – LBNL 57

Holographic models • Ad. S/CFT correspondence: strong coupling between jet and medium • The energy loss of a parton quenched in a thermal plasma is described by the dynamics of its dual classical string moving in a five-dimensional Ad. S space with a black hole April 17 th, 2013 Alessandro Buzzatti – LBNL 58